Chris Elrod
Chris Elrod
I agree. I've had trouble, too. [Here](https://github.com/chriselrod/LoopVectorization.jl/blob/cc71d9689470b0311184bf8b1372f58ab9a4d0a0/benchmark/plotbenchmarks.jl#L36) is the function I used for plotting benchmarks. I was using [category20](https://vega.github.io/vega/docs/schemes/) to define the colors. Maybe one of those is better than...
[Thoughts](https://chriselrod.github.io/LoopVectorization.jl/latest/examples/dot_product/)? I think Julia and ifort may look a little similar.
> One possibility is to use something like PlotlyJS where you can enable disable plots interactively, would make it easier to compare e.g. LoopVectorization vs vanilla Julia by turning off...
> I am many steps removed from Web development :p Me too. For now, I'd prefer a simpler solution. Or at least something that lets the most interesting contenders stand...
Hmm. That's a good idea. I think the current plots are a lot interesting to me personally, because I'm more interested in how their performance changes in absolute terms (e.g.,...
I'll add support for VectorizedRNG.jl.
manually provide reduction isomorphisms so that expensive operations can be factored out of the loop
Related: https://github.com/JuliaLinearAlgebra/Octavian.jl/issues/102
Julia lowers `NTuple{L,Core.VecElement{128}}` as LLVM arrays rather than as LLVM vectors, so it would not be trivial to add support.
The above commit disables threading when you have `CartesianIndices`. We can leave this issue open until we get a proper fix. The problem basically is that, given CartesianIndices, LV updates...
The easiest fix in case of `CartesianIndices` might be to call `condense_loopset` again, and take another trip through `_turbo_!` with the expanded loops.